LinkedIn Algorithm Claude Skill
The LinkedIn Algorithm Claude skill diagnoses weak organic reach and plans better posts for creators, founders and social teams. It is free and open source under the MIT license, from the LeadMagic gtm-skills repo. Its advice is built on Richard van der Blom's annual Algorithm Insights Report from Just Connecting.
/plugin marketplace add LeadMagic/gtm-skills/plugin install gtm-skills@gtm-skillsnpx skills add LeadMagic/gtm-skills --skill linkedin-algorithmWhat the LinkedIn Algorithm skill does
Many LinkedIn reach problems come from mechanics, not writing quality. External links in the post body, heavy hashtag use and too many tags can all hurt. This skill checks for those first, and separates a platform-wide decline from a personal one.
It starts from 90 days of your own post data, then reviews format mix, post anatomy and engagement habits. Format advice covers document posts, short video, image posts and text posts, using the ranking in the van der Blom report that the skill draws on.
It also covers the first-hour reply routine, saves as a signal, posting frequency, and how company pages and employee resharing fit in. Every recommendation is meant to trace back to the report rather than to folklore.
- Builds a 90-day baseline table from your post data
- Separates platform-wide reach compression from a decline specific to your account
- Flags reach killers such as external links, many hashtags and heavy tagging
- Recommends a format mix and when to use document posts, video, images or text
- Checks hooks, post length and keywords instead of hashtag reliance
- Sets up a first-hour reply routine and reciprocal commenting habits
- Advises on posting frequency and company page and employee resharing strategy
- Sets a 30-day remeasurement date using the same metrics
How it works
Pull a baseline
Claude asks for your last 90 days of posts with format, length, links, hashtags, timing and results, and puts them in one table.
Find reach killers
It separates platform-wide decline from your own, then flags external links, hashtag habits, heavy tagging and posting too often.
Fix the format mix
It ranks your formats against the report's findings and suggests where document posts, video or short text belong.
Shape each post
Hook, length and keywords get reviewed so the first lines hold attention and links move into comments or your profile.
Plan engagement and remeasure
You get a first-hour reply routine, a posting cadence, company page advice and a 30-day check on the same metrics.
Example prompts
Once the LinkedIn Algorithm skill is installed, ask Claude in plain language. These prompts trigger it:
- My LinkedIn impressions dropped by half this quarter. Diagnose what changed and what to fix
- Should I use a carousel or a video for this post about our pricing change?
- Do hashtags and external links still hurt LinkedIn reach?
- Build a posting cadence and first-hour engagement routine for our founder
- How should our company page post differently from our employees?
What you get
- A 90-day baseline table and reach-killer findings
- A format mix recommendation tied to the report
- A post anatomy checklist covering hook, length, keywords and tags
- A posting calendar and first-hour engagement routine
- A 30-day remeasurement plan
Who it is for
- Creators and founders whose LinkedIn reach has dropped
- Social media managers planning formats and cadence for a company page
- Marketers who want data-backed LinkedIn advice rather than tips from threads
How to install the LinkedIn Algorithm skill
What you need: Nothing beyond Claude and your own post analytics. The skill ships with a small Python checker for its output, which only matters if you run it in Claude Code.
- Claude Code plugin
/plugin marketplace add LeadMagic/gtm-skills - Claude Code plugin
/plugin install gtm-skills@gtm-skills - Skills CLI
npx skills add LeadMagic/gtm-skills --skill linkedin-algorithm - Manual
git clone https://github.com/LeadMagic/gtm-skills.git && cp -r gtm-skills/skills/inbound/linkedin-algorithm ~/.claude/skills/
After installing, restart Claude Code or start a new session. Claude loads the LinkedIn Algorithm skill automatically when your request matches it, or you can ask for it by name.
Tips for better results
- Give it real numbers from your last 90 days. Without a baseline, it can only give general advice.
- Test one change at a time, such as moving links to comments, so you can see what moved reach.
- Schedule the first-hour reply window before you publish, not after.
Limitations
- Its guidance rests on a third-party report, and LinkedIn can change how ranking works at any time.
- It cannot read your LinkedIn analytics directly. You supply the data.
LinkedIn Algorithm Claude skill FAQ
Yes. It is open source under the MIT license, from the LeadMagic gtm-skills repo. The Algorithm Insights Report it draws on is a separate third-party study.
Run /plugin marketplace add LeadMagic/gtm-skills, then /plugin install gtm-skills@gtm-skills. You can also use the Skills CLI line shown here, or copy the skill folder into ~/.claude/skills/.
The skill is mostly SKILL.md plus reference files, so it can be zipped and uploaded in Claude.ai under Settings, Capabilities, Skills on paid plans. The small Python output checker needs a terminal, so full use is in Claude Code.
It may be platform-wide, since the report this skill uses describes broad declines, or personal, from links in the body, hashtag habits or heavy tagging. The skill compares your 90-day data to tell them apart.
The skill says yes. It treats links in the post body as a common self-inflicted reach penalty and recommends putting them in the first comment or on your profile instead.